HEMOCHROMATOSIS GENOTYPES AND ELEVATED TRANSFERRIN SATURATION

Doctor of Medical Science Thesis by Christina Ellervik MD, PhD HEMOCHROMATOSIS GENOTYPES AND ELEVATED TRANSFERRIN SATURATION - risk of diabetes mell...
Author: Rose Blake
1 downloads 2 Views 1MB Size
Doctor of Medical Science Thesis by

Christina Ellervik MD, PhD

HEMOCHROMATOSIS GENOTYPES AND ELEVATED TRANSFERRIN SATURATION - risk of diabetes mellitus, hypertension,cancer, and total mortality

Affected Unaffected Carrier

Carrier Father

Affected Son

Carrier Mother

Carrier Daughter

Carrier Son

Unaffected Daughter

Hemochromatosis genotypes and elevated transferrin saturation - risk of diabetes mellitus, hypertension,cancer, and total mortality

Doctor of Medical Science Thesis by

Christina Ellervik PhD MD, PhD

The Faculty of Health and Medical Sciences at the University of Copenhagen has accepted this dissertation, which consists of the already published dissertations listed below, for public defence for the doctoral degree in medicine.

Copenhagen, October 11th 2015 Ulla M. Wewer Head of Faculty

Place and time for defence: St. Auditorium at Herlev Hospital, June 22nd 2016 at 2pm

Table of Contents • Papers on which the thesis is based............................................................. 2 • Preface ............................................................................................................. 3 • Scope and delimitation of the thesis ...................................................... 3 - 4 • Introduction ............................................................................................ 4 - 14 Hereditary hemochromatosis ............................................................................................ 4 - 7 Diabetes mellitus (paper 1 and 2) ...................................................................................... 7 - 8 Hypertension (paper 3)............................................................................................................. 8 Cancer (paper 4) .................................................................................................................... 8 - 9 Mortality (paper 5, 6, and 7) ............................................................................................. 9 - 10 Other diseases associated with hereditary hemochromatosis, but not part of the thesis ................................................................................................................... 10 - 11 Classification of diabetes mellitus ................................................................................ 11 - 13 Danish registries for ascertainment of diagnoses used in the thesis ........................... 13

• Objectives .............................................................................................. 13 - 14 • Methods ................................................................................................ 14 - 00 Study populations .................................................................... ....................................... 14 - 16 Genotyping ............................................................................................................................... 16 Measurement of transferrin saturation ...................................................................... 16 - 18 Ascertainment of diagnoses ................................................................................................. 18 Statistics ............................................................................................................................ 18 - 22

• Results ................................................................................................... 22 - 41 Total mortality and incident diabetes, hypertension, cancer, and ischemic heart disease in non-participants vs. participants .................................. 22 - 27 Risk of diabetes mellitus (paper 1 and 2) .................................................................... 28 - 29 Risk of hypertension (paper 3) ..................................................................................... 29 - 31 Risk of cancer (paper 4) .................................................................................................. 31 - 33 Risk of total mortality (paper 5, 6 and 7) .................................................................... 33 - 40 Results on hemochromatosis genotypes C282Y/H63D and C282Y/ wild type in paper 1, 3, 4, 6 and 7 ......................................................................................... 40 Key findings ...................................................................................................................... 40 - 41

• Discussion – methodological considerations .................................... 41 - 47 Errors in sampling ........................................................................................................... 41 - 44 Health examinations .............................................................................................................. 44 Other factors influencing study validity .................................................................... 44 - 45 Patient audit: classification of diabetes mellitus ..................................................... 45 - 46 Ascertainment and misclassification of diagnoses ................................................. 46 - 47

• Discussion – results in perspective .................................................... 47 - 62 Diabetes mellitus (paper 1 and 2) ................................................................................. 47 - 51 Hypertension (paper 3) .................................................................................................. 51 - 52 Cancer (paper 4) ............................................................................................................... 52 - 53 Total mortality (paper 5, 6, and 7) ................................................................................ 53 - 56 Discussion of hemochromatosis genotypes C282Y/H63D and C282Y/wild type (paper 1, 3, 4, 6 and 7)...................................................................... 56 -57 Screening for iron overload .......................................................................................... 57 - 58 Strengths and limitations ............................................................................................. 58 - 59 Biological plausibility ..................................................................................................... 59 - 60 Penetrance ....................................................................................................................... 60 - 61 Expressivity ..................................................................................................................... 61 - 62

• Future perspectives and unresolved questions .............................. 62 - 63 • Summary in English .................................................................................... 63 • Summary in Danish ............................................................................. 63 - 64 • Abbreviation list .......................................................................................... 64 • References ............................................................................................ 64 - 85

1 1

Papers on which the thesis is based (According to PubMed style)

1. Ellervik C, Mandrup-Poulsen T, Nordestgaard BG, Larsen LE, Appleyard M, Frandsen M, Petersen P, Schlichting P, Saermark T, Tybjærg-Hansen A, Birgens H. Prevalence of hereditary haemochromatosis in late-onset type 1 diabetes mellitus: a retrospective study. Lancet 2001; 358:1405-9. 2. Ellervik C, Mandrup-Poulsen T, Andersen HU, Tybjærg-Hansen A, Frandsen M, Birgens H, Nordestgaard BG. Elevated transferrin saturation and risk of diabetes: three population-based studies. Diabetes Care 2011; 34:2256-8. 3. Ellervik C, Tybjærg-Hansen A, Appleyard M, Ibsen H, Nordestgaard BG. Haemochromatosis genotype and iron overload: association with hypertension and left ventricular hypertrophy. J Intern Med 2010; 268:252-64 4. Ellervik C, Tybjærg-Hansen A, Nordestgaard BG. Risk of cancer by transferrin saturation levels and haemochromatosis genotype: population-based study and metaanalysis. J Intern Med 2012; 271:51-63. 5. Ellervik C, Tybjærg-Hansen A, Nordestgaard BG. Total mortality by transferrin saturation levels: two general population studies and a metaanalysis. Clin Chem 2011; 57:459-66. 6. Ellervik C, Andersen HU, Tybjærg-Hansen A, Frandsen M, Birgens H, Nordestgaard BG, Mandrup-Poulsen T. Total Mortality by elevated transferrin saturation in patients with diabetes. Diabetes Care 2013; 36: 2646-2654. 7. Ellervik C, Mandrup-Poulsen T, Tybjærg-Hansen A, Nordestgaard BG. Total and causespecific mortality by elevated transferrin saturation and hemochromatosis genotype in individuals with diabetes – two general population studies. Diabetes Care 2014; 37: 444-452.

2 2

Preface My interest in hemochromatosis began as a medical student, when I met Henrik Birgens, Thomas Mandrup-Poulsen, Børge Nordestgaard and Anne Tybjærg-Hansen. Since then, they have encouraged me to continue research within this field. Their contribution to my scientific education has been very valuable and has lead to a fruitful and rewarding collaboration during the years. The present thesis is based on studies carried out within this collaboration over the past 10 years. I owe thanks to all the collaborators for always constructive comments and criticism. I am also grateful to other co-authors of the presented works, from whom I have learnt much. The assistance of the technical staff and medical doctors at Steno Diabetes Centre (Sonja Bigand, Merete Frandsen, Henrik Ullits Andersen) and at the departments of Clinical Biochemistry (Hanne Dam) and Hematology at Herlev Hospital is greatly appreciated. I appreciate the generous and free working conditions created by the head of Clinical Biochemistry at Herlev Hospital, DMSc Niels Fogh-Andersen. I appreciate the hospital directors at Næstved and Nykøbing Falster Hospitals, Department of Clinical Biochemistry at Næstved Hospital, the Heads of Research at Næstved Hospital (Sten Boesby, Jan Kvetny) and at Region Sjælland (Steffen Groth, Knud Rasmussen) for generously supporting my research. I also owe thanks to the secretariat of the Copenhagen City Heart Study (Merete Appleyard, Jacob Marrot, Connie Haugaard), who have always been ready to assist in every conceivable way, and also to the participants and to the initiators of the Copenhagen City Heart Study and the Copenhagen General Population Study. Finally, I want to thank family and friends for their support, love, and always good mood, especially my husband, Ulrik Juul Christensen, and our children, Caroline Juul Ellervik and Elisabeth Juul Ellervik, my parents in law Bente and Flemming Christensen, and Maren Weischer.

Scope and delimitation of the thesis The scope of this thesis is on transferrin saturation, as a measure of iron overload, and the homozygous hemochromatosis genotype C282Y/C282Y for the association with diabetes, hypertension, cancer and total mortality; risk of other C282Y compound and heterozyogous genotypes will also be shortly discussed. Other measures of iron overload exist, e.g. ferritin, and other genetic variants exist for hereditary hemochromatosis as this is a genetically heterogeneous disease. However, elevated transferrin saturation ≥50% has a sensitivity of 75%, a specificity of 95%, and a positive predictive value of 3.5% for detecting HFE C282Y homozygotes1, whereas ferritin is elevated in many other diseases than hereditary hemochromatosis (e.g. inflammation, high alcohol intake, fatty liver disease)1, 2. Thus, transferrin saturation is usually the first screening test for hereditary hemochromatosis, followed by genetic testing and ferritin measurement3. The thesis focuses on adult white individuals of Danish descent, as hereditary hemochromatosis is primarily a disease in individuals of Northern European origin4. Also, the thesis focuses on Type 1 hereditary hemochromatosis3, but does not cover other types of hereditary hemochromatosis or secondary hemochromatosis. The thesis does not cover heart disease apart from hypertension, nor does it cover liver disease, arthritis, or pituitary hypogonadism, also features of hemochromatosis3. In a previous PhD thesis by Christina Ellervik in 2007, the focus was solely on hemochromatosis genotypes and risk of ischemic heart disease5, stroke6, and a meta-analysis7 of the existing literature at that time including 31 other disease end-points, but not including mortality. It was evident from the meta-analysis in the PhD thesis that most of the studies in the literature were cross-sectional and case-control studies. In contrast, the articles on which the present doctorial thesis is based are primarily follow-up studies. Also, the metaanalysis revealed areas of research that had not previously been investigated: association of late-onset type 1 diabetes with hemochromatosis genotypes (paper 1)8 and survival of

3 3

these patients (paper 6)9, risk of diabetes mellitus according to transferrin saturation in a follow-up study (paper 2)10, risk of hypertension according to transferrin saturation and hemochromatosis genotypes (paper 3)11, and risk of any cancer according to hemochromatosis genotypes in a follow-up study (paper 4)12. Hereditary hemochromatosis is a disease in individuals primarily of Northern European descent; the research questions on total mortality in the general population (paper 5)13 and in individuals with diabetes (paper 7)14 had been conducted previously in an American study15 and in an Australian study16 of mixed ethnicities, respectively, but never in a Northern European white Population. Thus, this thesis is independent of the previous PhD thesis, none of the papers in this thesis formed any part of the PhD thesis, and the articles on which this thesis is based add new research areas not previously covered. The genotypings used in the articles in this thesis were done in 1999 for the Copenhagen City Heart Study (papers 1-7) and the Steno Diabetes Center (paper 1 and 6), in 2001-2007 for a new cohort at the Steno Diabetes Center (paper 2 and 6), and in 2007 for patients from Losartan Intervention for Endpoint Genetic Sub-study (paper 3) and for Copenhagen General Population Study (paper 2, 3, 5 and 7). The same genotypings of some of the participants were also used for two papers5, 6 on ischemic heart disease and stroke in the PhD thesis. The transferrin saturation measurements we used in the articles in this thesis were done in 1999 for the Steno Diabetes Center (paper 1 and 6), in 2001-2007 for a new cohort at the Steno Diabetes Center (paper 2 and 6), in 2007 for the Copenhagen City Heart Study, and in 2003-2007 for the Copenhagen General Population Study (paper 2, 3, 5 and 7). The transferrin saturation measurements did not form part of any of the articles in the PhD thesis.

Introduction Hereditary hemochromatosis Hereditary hemochromatosis is an iron-overload disorder characterized by iron accumulation throughout life in various organs, such as the endocrine pancreas, heart, liver, pituitary gland, joints, and skin3, 17. Iron overload results in progressive tissue damage and organ failure, including diabetes mellitus, heart failure, liver cirrhosis, liver cancer, pituitary hypogonadism, arthritis, and bronze colour of the skin due to hyperpigmentation3, 17. However, several years of iron accumulation are needed before tissue damage leads to clinical signs of organ failure. The earliest detectable biochemical anomaly is increased transferrin saturation3, which generally represents an increased intestinal iron absorption; this may be followed by an increased ferritin concentration indicating accumulation of cellular iron3, 17. Non-hereditary hemochromatosis or secondary hemochromatosis may arise from chronic use of iron supplements, chronic liver disease, and especially chronic red blood cell (RBC) transfusion therapy for anemia in patients with ineffective erythropoiesis such as patients with beta-thalassemia major or myelodysplastic syndrome, or in patients with chronic hemolytic disorders or chronic bone marrow failure18. The complications associated with transfusional iron overload are in principle similar to hereditary hemochromatosis, but with earlier onset19, and require iron chelation therapy18. In 1996, the HFE gene was identified, with homozygosity for C282Y being responsible for 84% of hereditary hemochromatosis in Europeans and with compound heterozygosity for C282Y/H63D explaining an additional 4%20. The locus for HFE is on chromosome 6p21.320. A G −> A substitution at nucleotide 845 changes cysteine to tyrosine in codon 282 (C282Y), and a C−>G substitution at nucleotide 187 changes histidine to aspartate in codon 63 (H63D). The genotypes C282Y/C282Y, C282Y/H63D, C282Y/wild type and, in some studies, H63D/H63D, but not H63D/wild type, are associated with evidence of iron overload21, 22. The risk of developing the clinical hereditary hemochromatosis phenotype is debated23, 24, with penetrance estimates ranging from less than 1%23 to almost 30%24 for individuals with

4 4

Table 1. Comparative overview of genetics in hereditary hemochromatosis (HH). Classic HH

Juvenile HH

Juvenile HH

OMIM Classification

Type 1

Type 2, subtype A

Type 2, subtype B Type 3

Type 4

Gene name

HFE

HJV

HAMP

TfR2

SLC40A1

Gene locus

6p21.3

1q21

19q13.1

7q22.1

2q32

Gene product

HFE

Hemojuvelin

Hepcidin

Transferrin receptor 2

Ferroportin

Pattern of inheritance

Autosomal recessive

Autosomal recessive

Autosomal recessive

Autosomal recessive

Autosomal dominant

Predominant cell distribution of iron

Parenchymal

Parenchymal

Parenchymal

Parenchymal

Reticuloendothelial

Main organs accumulating iron

Liver, endocrine glands, heart

Liver, endocrine glands, heart

Liver, endocrine glands, heart

Liver, endocrine glands, heart

Liver, spleen

2nd or 3rd

2nd or 3rd

4th or 5th

4th or 5th

4th or 5th Decade of onset of symptomatic disease

TfR2-related HH

Ferroportinrelated HH

Table is modified from A. Pietrangelo NEJM 2004;350:2383-97. OMIM: Online Mendelian Inheritance in Man. HFE: High iron Fe HJV: Hemojuvelin HAMP: Hepcidin Antimicrobial Peptide TfR2: Transferrin receptor 2 gene SLC40A1: Solute Carrier Family 40 member A1

C282Y/C282Y and with the least penetrance in women and the highest penetrance in men. The reason for the lower penetrance in women is that women bleed during the reproductive period. Thus, onset in women is usually after the menopause, whereas in men onset is usually in the 30-50ies3. The autosomal recessive form of hereditary hemochromatosis with homozygosity for C282Y is also termed classic or Type 1 hereditary hemochromatosis; however, other genetic and phenotypic variants exist3 (Table 1). This thesis is focusing on classical hereditary hemochromatosis with homozygosity for C282Y/C282Y. Population studies have provided information on allele frequencies of C282Y and H63D25. There is a north to south gradient in Europe of the C282Y allele frequencies with the highest reported in Northern Europe and in Northern European emigrants (5-11%) and the lowest in Southern Europe (0-5%), and vice versa for H63D with allele frequencies of 1020% in the south and 5-10% in the north of Europe. It is believed that C282Y originated by chance in a single Celtic26, 27 or Viking28 ancestor in North Western Europe about 2000-4000 years ago3, 25, and a joint hypothesis is that C282Y originated in the Celts but was spread by the Vikings29. The reasons why the allele frequencies of C282Y are higher in Northern Europe than in Southern Europe can be explained both by the founder effect, but it may also be speculated that heterozygotes have a survival advantage or that heterozygotes are better protected against anemia30, which could be beneficial in the reproductive period for younger women; however, there is no difference in parity status among heterozygous women compared to wild types30. As will also be seen later in this thesis in the Results section, risk of mortality in individuals with C282Y/wild type is not different from that in wild type/wild type individuals; however, mortality as studied here in later life will not affect reproduction. Most individuals homozygous for C282Y have elevated levels of transferrin saturation and ferritin23, 31-34. Biochemically, cut-off values for transferrin saturation and ferritin in the diagnosis of hemochromatosis have varied across studies35, but a transferrin saturation, reflecting plasma iron, above 50% and a ferritin concentration, reflecting tissue iron, above 1000 μg/L are generally accepted criteria, although there are gender differences35.

5 5

Figure 1.

Hepcidin and iron influx.

The normal infux of iron from enterocytes to the blood stream is mediated through the transporter ferroportin on the basolateral side of the enterocyte. The hormone hepcidin, which is secreted by hepatocytes, is the central regulator of iron homeostasis. Hepcidin transcription is upregulated in response to iron overload and downregulated in response to iron deficiency. In conditions with iron overload, hepcidin downregulates the ferroportin-mediated release of iron from enterocytes and macrophages to the blood. Except for inherited defects of hepcidin itself (HAMP) and ferroportin (SLC40A1), all forms of iron-storage disease appear to arise from hepcidin dysregulation. Hepcidin is regulated by various proteins, such as HFE (HFE) and transferrin receptor 2 (TfR2), and mutations in the hemojuvelin gene (HJV). See also Table 1.

Duodenal lumen

Liver

Iron

HJV

Regulation

Dimetalion transporter

HAMP

Enterocyte

Hepcidin

Ferritin

ul at io

n

Iron

HFE Protein

Re g

Ferroportin (Iron transporter) Basolateral membrane

TRf2

Re g

ula tio

n

HFE

Iron

Hepcidin

Transferrin receptor 1

Plasma

Transferrin bound iron

There is now substantial evidence that the liver plays a central role in determining, how much iron is absorbed from the intestinal tract and in influencing the release of iron from sites of storage36. The normal flux of iron from enterocytes to the blood stream is mediated through the transporter ferroportin on the basolateral side of the enterocyte (Figure 1), while the hormone hepcidin, which is secreted by hepatocytes, is the central regulator of iron homeostasis. Hepcidin transcription is up-regulated in response to iron overload and down-regulated in response to iron deficiency37. In conditions with iron overload, hepcidin down regulates the ferroportin-mediated release of iron from enterocytes and macrophages to the blood36. Iron-storage disease may arise due to mutations in the genes coding for the HFE protein (HFE; Type 1), hemojuvelin (HJV; Type 2A)38, hepcidin( HAMP; Type 2B), transferrin receptor 2(TfR2; Type3), and ferroportin (SCL40A1; Type 4)3, 36 (Table 1). The underlying mechanism of action for the iron-induced cellular damage involves formation of free radicals through the Fenton reaction resulting in oxidative stress to the cells39, 40. Depending on which cells are damaged, different diseases in different organs develop. Clinically, hereditary hemochromatosis is thought of as rare41, 42; however, genotyping

6 6

studies suggest that it potentially is one of the most common genetic diseases in people of Northern European descent4, since the prevalence of the genotype C282Y/C282Y is 0.25% and the prevalence of the genotype C282Y/wild type is 9%8 .

Diabetes mellitus (paper 1 and 2) In vitro rat beta-cells are prone to reactive oxygen species (ROS)-mediated damage by cytokine-dependent up-regulation of cellular iron import by the divalent metal ion transporter (DMT-1)43, and inducible beta-cell specific knock-out of DMT-1 protects against inflammatory and high fat diet induced diabetes. Also, introduction of HFE mutations in mice results in beta-cell oxidative stress and decreased insulin secretory capacity due to glucose desensitization and to beta-cell apoptosis44. Further, increased hepatic glucose production and decreased skeletal muscle glucose oxidation may contribute to diabetes in iron overload in an animal model of Type 2 diabetes45. Accordingly, iron restriction or iron chelation protects from diabetes and loss of beta-cell function in obese mice46. The defects in both insulin-producing and insulin sensitive tissues are most likely caused by iron-dependent catalysis via the Fenton reaction of ROS, which impair insulin signalling in skeletal muscle and liver, and which cause beta-cell destruction due to insufficient betacell antioxidant defence47, 48. In autopsies of patients with hereditary or non-hereditary hemochromatosis and diabetes, impaired glucose tolerance, or impaired fasting glucose, iron was found deposited in the beta-cells49-54. Also, there is an inverse relationship between beta-cell iron deposits and insulin secretory granules50, 55, and the number of beta-cells with hemosiderin increases proportionally with blood transfusion volume51. Furthermore, the degree of impaired fasting glucose and glucosuria increases with increasing iron overload and with increasing beta-cell iron staining51. In patients with only mild hemosiderosis of the pancreas iron staining is confined to acinar cells and the stroma51, 53, and control autopsies of patients without a history of hemochromatosis, blood transfusions or diabetes do not show any iron staining in the pancreas51. In pancreatic tissue from autopsies of patients with hemochromatosis the histological appearance of the islets is normal with their shape and size being unchanged (unlike Type 1 diabetes)49, amyloid deposits are absent (unlike Type 2 diabetes) 49, 53, and there is no iron staining in the alpha or delta cells49-52. Thus, pancreatic involvement in iron overload seems to be a continuum from mild hemosiderosis involving acinar cells and the stroma to severe hemosiderosis with beta-cell involvement. Functionally, diabetes secondary to hereditary or non-hereditary hemochromatosis is characterised by both insulin resistance56-60 and insulin deficiency48, and may therefore mimic both Type 2 diabetes and idiopathic Type 1 diabetes. Patients with hereditary hemochromatosis are more prone to develop diabetes if insulin resistant, since their insulin secretory capacity is decreased but usually not absent48, 61. In 2001 (paper 1), we postulated that the most likely explanation for the discrepancy that hemochromatosis was thought of as rare but the frequency of the HFE C282Y homozygosity was 0.25% with an estimated penetrance rate of 50%, was that hereditary hemochromatosis was often overlooked8. Our aim was to test this hypothesis in patients with lateonset Type 1 diabetes mellitus, a hypothesis that had not been investigated before at that time. Previous studies of risk of diabetes conferred by the C282Y/C282Y genotype had primarily focused on Type 2 diabetes or any diabetes with contradicting results7, 23, 62-82. The evidence now in 2014 from epidemiological studies is that hereditary or non-hereditary hemochromatosis is associated with increased risk of developing or dying from diabetes7-10, 14, 83-86. It is also now known that high serum ferritin correlates inversely with serum insulin, and proportionally with blood glucose concentrations87. Also, high serum ferritin, body iron, haem intake, and dietary iron intake are associated with increased risk of diabetes83-86, 88-91. Finally, phlebotomy may improve insulin secretory capacity 92, 93 and insulin sensitivity94 if instituted early, and it reverses impaired glucose tolerance in patients with hereditary hemochromatosis93, 95.

7 7

In the context of the existing literature, three studies86, 96, 97 have used transferrin saturation as the primary independent variable (without considering hemochromatosis genotype) for the study of the risk of any diabetes. However, these studies are heterogeneous with regard to ethnicity, study design, size, transferrin saturation threshold, diabetes diagnosis, source of diabetes diagnosis, and conclusions as to whether elevated transferrin saturation is associated with risk of diabetes (one study in favour96 and two studies not in favour86, 97). Therefore, we undertook another study (paper 2) of whether elevated transferrin saturation conferred increased risk of any type of diabetes mellitus, type 1 diabetes mellitus, and type 2 diabetes mellitus in three independent Danish white population-based studies10 from the same geographical area.

Hypertension (paper 3) and other heart disease (cardiomyopathy, ischemic heart disease, cardiac arrhythmias) In asymptomatic subjects with C282Y/C282Y genotype there appears to be an early detectable echocardiographic manifestation of abnormal diastolic function98, which may be associated with oxidative stress caused by the iron overload99, but systolic function seems to be preserved100. The echocardiographic features of patients with symptomatic hemochromatosis are varying degrees of cardiomyopathy and left ventricular hypertrophy (LVH)101, the latter also a feature of hypertension102. Increased thickness of the ventricular wall is probably the first, and still reversible, cardiac alteration due to iron deposition in the myocardium103. Later, with increasing iron overload, left ventricular function becomes impaired and dilated cardiomyopathy develops103. In support, accumulating evidence suggests that oxidative stress may alter the modulation of vascular tone104, 105, thereby affecting blood pressure leading to hypertension. Furthermore, it has been shown that arterial wall thickness was increased before the onset of cardiovascular complications in hemochromatosis patients and that this alteration was reversed by iron depletion106. Also, in men with hypertension increased serum ferritin has been shown to be more frequent than in controls107. It is therefore possible that individuals with hemochromatosis genotypes and⁄ or iron overload are overrepresented amongst patients with hypertension and⁄or LVH11, a hypothesis that was tested in paper 3. Meta-analyses have not shown any association of iron overload 108 or HFE genotypes7, 109 with ischemic heart disease or myocardial infarction; also, HFE genotypes are not associated with oxidized LDL5. In single case-study of patients with symptomatic hereditary hemochromatosis, severe cardiac arrhythmias have been described110, 111; however, in a larger case-study of symptomatic patients with hereditary hemochromatosis, ECG abnormalities were non-specific compared to controls103. In asymptomatic patients with C282Y/C282Y genotype self-reported arrhythmias were not different from that in wild-type subjects112, but ECG-measurements suggested a marginal but significant difference in non-life-threatening ECG abnormalities113. However, it is estimated that in male patients with pacemaker treated atrioventricular block of second or third degree, that the prevalence of patients with hereditary hemochromatosis is 1.3%114, i.e. 4-5 times the prevalence of C282Y homozygosity in the general population.

Cancer (paper 4) Iron-induced free radical damage to DNA may be important for the development of cancer115, and cancer cells grow rapidly in response to iron116, 117. Thus, iron overload, either in general or genetically via hemochromatosis genotype C282Y/C282Y, may lead to increased risk of cancer. The most severe outcome of hereditary hemochromatosis is liver cancer3, also associated with hemochromatosis genotype C282Y/C282Y in case-control studies7. However, a meta-analysis of the association of any other cancer than liver cancer with hemochromatosis genotype C282Y/C282Y in case-control studies had insufficient power to exclude an association7. Risk of cancer in individuals with iron overload has previously been studied in

8 8

various prospective118-129 and case-control studies130, but with contradicting results. However, it has been shown that therapeutic iron reduction in a randomized trial131 and blood donation132 both seem to reduce cancer risk. No population-based follow-up study of cancer risk as a function of hemochromatosis genotypes has previously been conducted. Therefore, we recently conducted such a study along with a meta-analysis of elevated transferrin saturation and risk of any cancer12 in paper 4.

Mortality (paper 5, 6 and 7) Paper5

The 4th and 5th decade is usually the onset of symptomatic organ disease in individuals with hereditary hemochromatosis, with onset for men not uncommon in their thirties and for women after menopause3. There is even evidence for increased mortality in patients with clinically overt hereditary hemochromatosis133-137. Early diagnosis and instigation of appropriate treatment with repeated venesections can prevent the consequences of hereditary hemochromatosis and restore normal life expectancy134, 138, 139. Total mortality according to increased transferrin saturation, a biochemical proxy for iron overload, was examined in a single previous population-based study (NHANES)15; this study examined 10,000 individuals including 10% non-whites, who potentially could have a very low chance of having iron overload, since iron overload is a disease most often seen in individuals of Northern European origin140. Furthermore, this study did not examine a dose-response relationship from low iron loads to iron overload, did not examine extreme phenotype (iron overload above 70%) but only above 45-60%, and did not show gender-stratified results. Due to these weaknesses in the NHANES study and since significant results can either arise by chance, confounding, bias, misclassification, or reflect a true association, we therefore conducted a similar study13 (paper 5). However, our study was larger (45,000 individuals, all whites of Northern European Danish descent); and as a novel idea and finding, in this paper, as was not presented in the NHANES paper, we added a semi-graded relationship of transferrin saturation(%) with cut-points of

Suggest Documents